Elaboration of the protocol
This protocol was developed according to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guideline [17] (Additional File 1).and the results will be reported according to the Consolidated Standards of Reporting Trials (CONSORT) statement when the study is completed [18].
Study design {8}
This is a prospective, parallel group, two-arm, superiority randomised controlled trial with 1:1 allocation ratio
Settings and eligibility criteria {9, 10}
Study participants will be people aged 18 to 60 years old seeking care for a new episode of acute ankle sprain at the emergency service of the Nossa Senhora da Saúde Hospital, in Diamantina, Brazil.
The inclusion criteria will be:
- People aged from 18 to 60 years old;
- Clinical diagnosis of grade I or II ankle sprains by a trained clinician, suggesting an incomplete ligament rupture [19];
- Duration up to 72 hours from the new episode of ankle sprain to the day of the medical appointment;
- Bone fractures excluded by the Ottawa ankle rules and radiography [20].
The exclusion criteria will be as follows:
- Grade III (severe) ankle sprain, suggesting complete ligament injury, determined by a clear positive test of the anterior drawer and / or inversion stress test, accompanied by severe swelling, haemorrhage, high level of pain on palpation, in addition to total loss of the ability to support weight on the foot and of the dorsiflexion range of motion [19].
- Open injuries that contraindicate the application of ice (e.g., any degree of vascular ulcer);
- Application of any cryotherapy more than once since the episode, before the allocation process;
- Application of any cryotherapy after being assigned to control group;
- Any condition that contraindicate the application of ice (e.g., Raynaud's syndrome), or any other intervention prescribed in the trial.
Intervention {11a, 11b, 11c, 11d}
The intervention will be reported according to the Template for Intervention Description and Replication (TIDieR) checklist and guide [21] (Additional File 2).
Ice Group {22}
Participants allocated to the ‘Ice Group’ will receive home prescription to submerge the ankle on a bucket of ice and water until cover the area of swelling and pain, sitting with 90° knee flexion in the affected limb, for up to three times per day, for 20 minutes, during seven days. In addition to cryotherapy, ankle elevation above the chest level, NSAID (i.e., nimesulide 100 mg, 2 times a day, during 5 days) and medical advice to rest for three days are prescribed [22, 12]. Daily phone calls and/or text messages by physiotherapists motivate record of applications using an intervention diary to assess adherence and adverse effects (Additional file 3). When participants do not adhere to immersion, alternative ice packs are prescribed in the same dosage.
Evidence from basic research suggests that cryotherapy may act to reduce pain intensity, inflammation and swelling, leading to improvement on ankle function. In pain, cryotherapy can decrease nerve conduction velocity, decreasing muscle spasms generated by spinal reflex after trauma [23], in addition to stimulating thermoreceptors that could inhibit the processing of nociception signals by the central nervous system, increasing the pain threshold [14]. By decreasing the local temperature, cryotherapy can reduce the metabolic demand of the area, avoiding the formation of swelling and the risk of a secondary injury caused by post-traumatic hypoxia and consequently preventing cell death [24]. Other mechanisms that can contribute to the reduction of swelling are local vasoconstriction, which leads to a decrease in blood flow to the tissue and a decrease in vascular permeability [25, 26].
No Ice Group
Participants allocated to the ‘No Ice Group’ will receive the same interventions of the ‘Ice group’ but with no ice. The prescription will be consisted of ankle elevation above the chest level, NSAID (i.e., nimesulide 100 mg, 2 times a day, during 5 days) and medical advice to rest for three days. Daily phone calls and/or text messages will also motivate record of applications in the ‘No Ice Group’.
Patients will be advised to not be part of any rehabilitation program until the end of the trial (i.e., 12 weeks after allocation). In case of non-compliance with this orientation, we will perform sensitivity analysis to minimise the concomitant care effects. Adverse effects have rarely been reported in the literature. In case of superficial skin burn, itching and cold-induced peroneal nerve palsy or any adverse symptoms that may be related to the interventions in this study, the patient will be instructed by the physiotherapist to discontinue their assigned interventions. They will be analysed normally according to their allocation in all segments and will be reported as an adverse event.
Outcome measures {12}
Primary outcomes
The primary outcome will be function, measured with the 0-80 Lower Extremity Functional Scale (LEFS) questionnaire, with higher scores meaning better functional status, and a Minimum Clinically Important Difference (MCID) of 9 points [27].
Secondary outcomes
The average pain intensity in the last 24 hours, measured with an 11-point Numerical Rating Scale (NRS), with higher scores meaning worse pain intensity [28] and MCID of 1.3 points [29];
Swelling, measured using the figure-of-eight method, which consists of circumference using a tape measure of the areas with the highest concentration of ankle swelling (the region of the anterior talofibular, calcaneofibular and anterior tibiofibular ligaments). The measurement is made by positioning the starting point (0) of the tape measure over the midpoint between the articular projection of the tibialis anterior tendon and the lateral malleolus, directing the tape to the centre of the medial longitudinal arch of the foot, over the navicular bone , passing through the base of the fifth metatarsal and crossing the upper face of the midfoot towards the lower point of the medial malleolus, passing through the calcaneal tendon, lower point of the lateral malleolus, until finding the zero point of the measure tape [30]. The Minimum Detectable Change (MDC) is 0.96 centimetres (cm) [31];
Dorsiflexion range of motion, assessed by two measurements of active ankle goniometry with the patient in the prone position (knee extension and with 90° of knee flexion to reduce the influence of the gastrocnemius). The ankle will be positioned in a position of 0° of movement in the sagittal plane, placing the goniometer axis over the calcaneus, with the fixed arm aligned with the calcaneal tendon and the mobile arm aligned with the second metatarsal. A digital goniometer will be used as measurement tool. The participant will be instructed to perform as much dorsiflexion as possible [32, 33, 34]. The MDC for this measurement is 6 degrees (o) [32].
Assessors were trained to perform the dorsiflexion range of motion goniometry and the figure-of-eight technique followed by a pilot study for intra- and inter-examiner reliability. Data for the Intra-class Correlation Coefficient (ICC) and its 95% confidence interval (95%CI) were collected on two different measurement occasions, with an interval of 1 week. We recruited six individuals of both sexes (three males and three females), collecting measurements from both lower limbs (n =12). We found a ICC of 0.95 (95% CI 0.80 to 0.98) for intra-examiner and 0.93 (95% CI 0.83 to 0.98) for inter-examiner reliability for goniometry. For the figure-of-eight technique, a ICC of 0.97 (95% CI 0.84 to 0.99) for intra-examiner and 0.96 (95%CI 0.88 to 0.99) was found.
Procedures and participant timeline {13, 26a}
All eligible participants will be informed about the study and must sign the consent form provided by the physicians prior to participation (Additional File 4). Baseline assessment will include age, Body Mass Index (BMI), sex, dominant limb, history of previous ankle sprains, ability to bear weight on the affected ankle (Yes/No), comorbidities. The degree of injury will be classified according to the Birrer et al. (1999) [19] classification. The outcomes of interest (function, pain intensity, swelling and dorsiflexion range of motion) will be collected at the baseline and reassessed at the following time-points: Short term (i.e., 7 to 14 days after allocation), and long-term (i.e. 12 weeks after allocation). Furthermore, we will investigate the immediate effects of cryotherapy on pain intensity in a time-point between 24h to 48h after baseline. A schematic diagram is available in Figure 1.
Figure 1 – Schematic diagram of the procedures and participant timeline. t1 – Baseline assessment; t2 – 24-48assessment (immediate effects); t3 – 7-14 days’ assessment (short-term; t4 – 12 weeks’ assessment (long-term).
Data analysis
Sample calculation {14}
Function (LEFS) – primary outcome
The sample calculation was performed considering the MCID value of 9 points for the measurement of the primary outcome [27], and a standard deviation of ±12.85 based on a previous study [4]. A sample of 82 participants (41 per group) is necessary for a minimum detection of the effect size, taking into account a statistical power of 80%, α of 5% and a dropout rate of 20%.
Pain Intensity (0-10 NRS) – secondary outcome
The sample calculation was performed considering the MCID value of 1.3 points [29], and a standard deviation of ±0.8 based on a previous study [35]. A sample of 16 participants (8 per group) is necessary for a minimum detection of the effect size, taking into account a statistical power of 80%, α of 5% and a dropout rate of 20%.
Swelling (figure-of-eight method) - secondary outcome
The sample calculation was performed considering a difference of 3.4 cm and a standard deviation of ±3.8 based on a previous study [31]. A sample of 50 participants (25 per group) is necessary for a minimum detection of the effect size, taking into account a statistical power of 80%, α of 5% and a dropout rate of 20%.
Dorsiflexion range of motion (Goniometry) - secondary outcome
The sample calculation was performed considering a difference of 6° and a standard deviation of ±6.8 based on a previous study [35]. A sample of 52 participants (26 per group) is necessary for a minimum detection of the effect size, taking into account a statistical power of 80%, α of 5% and a dropout rate of 20%.
Recruitment {15}
Eligible people seeking care for acute ankle sprain at the emergency of the Nossa Senhora da Saúde Hospital will be referred to researchers who will contact them to check eligibility, explain the trial and ask for formal consent. To inform the population about the ongoing trial, social media and posters will be used.
Randomisation e allocation {16a, 16b, 16c}
The randomisation sequence for the experimental and control groups, with a 1:1 allocation ratio, will be computer-generated by one of the researchers who will not be involved in recruiting and assignment. The sequence will be generated in blocks of 4, 6 and 8, in random order. Doctors working in the emergency department will enrol and assign participants to the groups. The allocation will be hidden in sequentially numbered opaque sealed envelopes. All procedure will be conducted following recommended methods [36].
Blinding {17a, 17b}
The statistician will be blinded to the allocation of the participants. The data will be encoded in an unidentifiable way and will not contain any information that can raise suspicions about the allocation of the participants. Additionally, the assessors will be blinded for the secondary outcomes (i.e., figure-of-eight method and goniometry). The circumstances in which unbliding is allowed will be if the application of ice or any of the interventions applied in this study generate serious side effects or damage that requires medical assistance.
Analysis of treatment effects {20a, 20b}
The statistical analysis will be performed following the intention-to-treat principle. The normality of the data will be tested using the Kolmogorov-Smirnov test and the homoscedasticity of the data will be tested using the Levene test. Parametric data will be expressed as means and standard deviation and analysed with Mixed-Effects Models for repeated measures considering group versus time interaction and treatment effects adjusted for baseline values and a post-hoc Bonferroni analysis. In cases of non-parametric data, median and its upper and lower limits will be expressed. and analysed using the Generalized linear mixed effects models. All statistical analyses will be performed using the Stata software version 17.0 (StataCorp LLC, College Station, United States). The effect sizes will be interpreted based on MCIDs of our outcomes of interest. There are no subgroup analyses planned.